Interview, Podcast
How Writing Led To Investing In Ramp, Scale AI, Replit, Varda | Packy McCormick, Not Boring
- Fund Growth & AUM: The Not Bored Capital fund has grown to approximately $50M in Assets Under Management (AUM) after successfully raising a third fund despite a difficult fundraising environment during the market crash.
- Investment Thesis Shift: The strategy pivoted toward "vertically integrated" companies combining hardware and software to outperform incumbents on speed, cost, and quality, explicitly rejecting pure software bets in favor of "bits and atoms" infrastructure.
- Portfolio Composition: The fund utilizes a "barbell" approach: 80% in vertically integrated systems (e.g., Earth AI, Base, Fuse) and 20% in high-risk, high-reward sectors like crypto and deep tech where the founder's insight is paramount.
- Early Revenue Trigger: The newsletter's monetization began roughly nine months after launch when a single $1,000 ad deal occurred after the creator tweeted an ad deck, leading to a full-time transition from his previous job at the end of 2019.
- Current Subscriber Base: The "Not Boring" newsletter has expanded from 400 initial subscribers to over 239,000 subscribers, evolving from a newsletter into a venture fund and media company.
- AI Skepticism: The host argues AI will become an "incredibly profitable super SaaS" rather than an existential threat ("God or Satan"), noting no qualitative "leap" to sentience has been observed in recent model releases like DeepSeek or GPT-5.
- Commoditization of Intelligence: AI models are viewed as a commoditized layer where price and brand differentiation matter most; the host predicts a shift toward brand loyalty similar to the iPhone, where users pay a premium for trusted "intelligence."
- DeepSeek Impact: The release of DeepSeek caused a market correction in the AI narrative, revealing that significant algorithmic efficiency gains (doing more cheaper) are replacing the narrative of massive compute scaling as the primary value driver.
- Vertical Integration Case Study (Base): Base Energy, founded by ex-Anduril and Thrive/Blackstone executives, is investing in residential battery storage to solve grid transmission bottlenecks, operating as a retail power provider that rents batteries to customers in Texas.
- Base Financial Model: Base avoids the traditional battery manufacturing deep tech risk by assembling packs, while generating recurring revenue through energy trading and installation fees, leveraging a partnership with homebuilder Lennar for distribution.
- Vertical Integration Case Study (Earth AI): The mining company Earth AI utilizes a vertically integrated model to sell stakes in specific mines for immediate cash flow, funding expansion without needing traditional external venture capital rounds.
- Vertical Integration Case Study (Fuse): Fuse Power, led by founder JC, aims to become the "SpaceX of the National Nuclear Safety Administration," selling pulse power generators for nuclear testing and rad-hard electronics as a cash-flow engine toward fusion.
- Founder Evaluation Criteria: The primary metric for investment is the founder's ability to inspire exceptional work and their deep, specific knowledge of industry history and failure points, which serves as a "compression algorithm" for due diligence.
- Diligence Methodology: Investment decisions are often made within 30-60 minutes of meeting a founder based on intuition and fit, with the subsequent writing process serving as a tool to explain the thesis to the world and future investors rather than to make the initial decision.
- Capital Access for Hard Tech: While early-stage rounds (Series A) can be less competitive for hard tech, the "valley of death" exists at manufacturing scale-up; however, once a specific technical milestone is proven, later-stage capital (Series B/C) flows rapidly due to reduced execution risk.
- Nuclear & Defense Trends: The author notes a resurgence in defense and nuclear startups (e.g., SMRs) driven by the "American dynamism" trend and a shift toward viewing nuclear as a brand-driven category where "first mover" status creates a moat similar to SpaceX.
- YC Critique: The host expresses skepticism regarding Y Combinator's current startup requests, believing that pre-identified categories often attract founders with "confirmation bias" rather than genuine obsessions with solving specific problems.
- Writing as Diligence: The author views deep-dive journalism (sometimes 90+ pages) as the superior form of early-stage underwriting, arguing that understanding the "room with Lyndon Johnson" via narrative is more valuable than market maps or financial projections alone.
- Future Outlook: The host remains optimistic about the future but expects AI to drive efficiency in vertically integrated companies rather than replace human decision-making, with the next decade focused on "bits and atoms" infrastructure rather than pure software.